The value of dynamic FDG PET/CT in the differential diagnosis of lung cancer and predicting EGFR mutations

Abstract Introduction: This study aims to evaluate the value of the quantitative metabolic parameters derived from dynamic 18 F-fluorodeoxyglucose (FDG) positron emission tomography/CT (PET/CT) in the differential diagnosis of lung cancer and predicting epidermal growth factor receptor (EGFR) mutation status. Methods: We included 147 patients with lung lesions who were enrolled to perform FDG PET/CT dynamic + static imaging with informed consent. Based on the results of puncture and postoperative pathology, the patients were divided into benign and malignant groups, adenocarcinoma (AC) and squamous carcinoma (SCC) groups and EGFR-positive (EGFR+) and EGFR-negative (EGFR-) groups. Quantitative parameters including K 1 , k 2 , k 3, and K i of each lesion were obtained by applying the irreversible two-tissue compartment modeling using in-house Matlab software. The standardized uptake values (SUV) analysis from conventional static data. Differences in each metabolic parameter among the groups were analyzed. Wilcoxon rank-sum test or Independent-samples T-test and receiver-operating characteristic (ROC) analyses were performed on each parameter to compare the diagnostic effects among the differentiated group. P< 0.05 were considered statistically significant for all tests. Results: In the malignant group ( N =124), the SUV max , k 2 , k 3 , and K i were higher than the benign group ( N =23), and all have good performance in the differential diagnosis ( P <0.05, respectively). In the AC group ( N =88) , the SUV max , k 3 , and K i were lower than in the SCC group, and the differences were statistically significant ( P <0.05, respectively). For ROC analysis, when the K i cut-off value of 0.0250 ml/g/min have better diagnostic specificity than SUV max (0.999 vs 0.70). In AC group, 48 patients underwent EGFR testing. In the EGFR (+) group ( N =31 ), the average K i (0.0279±0.0153 ml/g/min) was lower than EGFR (-) group ( N =17, 0.0405±0.0199 ml/g/min), and the differences were statistically significant ( P <0.05). However, the SUV max , and k 3 did not show such a difference between EGFR (+) and EGFR (-) groups ( P ＞0.05, respectively). For ROC analysis, the K i had a cut-off value of 0.0350 ml/g/min for predicting EGFR status, a sensitivity of 0.710, a specificity of 0.588, and an AUC of 0.674 [0.523-0.802]. Conclusion: When the cut-off value of K i was 0.0250 ml/g/min, there was a more specificity than SUV max for the differential diagnosis of lung cancer, although both methods were specific. The K i has a good diagnostic value in the prediction of the EGFR status. For patients for whom EGFR testing is not available, dynamic imaging may become an important non-invasive screening tool.